How to use moving averages: three setups and when they fail

Updated November 2026

Updated November 2026

How to use moving averages comes down to three setups: a trend filter, a crossover and a pullback level. Each one has a market that breaks it, and each false signal costs a spread or a fee.

Moving average trading is popular because the lines are simple to read. They are also late by design. This guide takes the three setups in turn, as one part of the main technical analysis indicators, and prices what goes wrong with each. On Ouinex you apply these averages to the charts of CFDs and perpetuals, and speculate on price without owning the asset.

CFD and crypto perpetual trading involves risk of loss. Leverage amplifies both gains and losses, and every false signal costs a spread or fee whether the trade wins or not.

How to use moving averages: the short answer

Traders use moving averages in three main ways: as a trend filter, as a crossover signal and as a level to watch on pullbacks. All three lag price, and all three fail in sideways markets.

They lag because a moving average is built from past prices, and the longer the period, the greater the lag. The table sums up the three setups covered below.

Setup	How traders use it	What breaks it	What a false signal costs
1. Trend filter	Price above a long average is read as an uptrend, below it as a downtrend	Sideways markets: price crosses the average again and again	A full round trip each time a flip triggers a trade
2. Crossover	A faster average crossing a slower one is read as a change in direction	Lag, and ranges where crosses reverse within days	8 false crosses in a month: worked out below
3. Pullback to the average	In a trend, a short average is watched as a level where pullbacks have paused	The trend ends, or a gap skips the level	Slippage on the stop, plus the round trip

SMA or EMA, and which length?

A simple moving average (SMA) weights every close in its window equally. An exponential moving average (EMA) weights recent prices more, so it turns sooner, and for the same reason it whipsaws a little more. A separate comparison of SMA vs EMA covers the difference in detail.

There is no general answer to which moving average is best, only conventions. Common lengths, listed as conventions rather than recommendations:

9 or 20 periods: often EMAs on intraday charts.
50 periods: medium term, often one side of a crossover.
200 periods: a long-term filter on daily charts.

Setup 1: the trend filter, and when it fails

How traders use it: price above a long average, often the 200-period on a daily chart, is read as an uptrend; below it, as a downtrend. Some traders only take trades in the direction the filter shows.
What breaks it: a sideways market. Price crosses the average again and again, so the "trend" flips with no trend at all. Traders call this a whipsaw.
What a false signal costs: each flip that triggers a trade is a full round trip. On a spread-priced market you pay the spread; on a crypto perpetual you pay the fee to open and again to close, plus the spread.
What to check first: whether the average itself is sloping or flat. A flat 200-period line with price weaving around it is the condition this setup fails in.

Setup 2: the moving average crossover, and when it fails

How traders use it: a faster average crossing a slower one, for example a 20-period over a 50-period, is read as a change in direction. In a moving average crossover strategy the cross itself is the trigger: the fast line crossing up is read as a long signal, crossing down as a short one.
What breaks it: lag first. By the time two averages cross, much of the move has often happened. Then ranges: in a sideways market, crosses come back within days.
What a false signal costs: a round trip per cross, worked out below.
What to check first: how far price has already moved since the turn began, and whether recent crosses in this market reversed quickly.

The cost of 8 false crosses in a month (illustrative)

Assume a sideways month produces 8 crosses that each reverse, and each one is traded as a round trip. The count of 8 is an assumption for the arithmetic, not a measured rate.

EUR/USD CFD, 10,000 euros (0.1 lot): on EUR/USD, a position of 10,000 euros moves $1 for each pip. Ouinex charges zero commission on forex, stock indices and commodities: you pay the spread. A round trip crosses the spread once, so 8 false crosses cost 8 × the spread in pips × $1. At an illustrative spread of 1 pip, that is $8; check the live spread on the platform before you count.
BTC perpetual, $6,000 notional: crypto perpetuals are 0.045%. That is $2.70 to open and $2.70 to close, or $5.40 a round trip, so 8 round trips cost 8 × $5.40 = $43.20, plus the spread, with the fee charged on both the opening and the closing trade in this example.

Those costs are paid whether each trade ends up ahead or not. Positions held overnight may also pay financing on CFDs or funding on perpetuals. Check Ouinex's trading fees for the market you trade before you count.

Setup 3: the pullback to the average, and when it fails

How traders use it: in a trend, some traders watch a short average, often a 20-period EMA, as a level where pullbacks have paused before. They look for price to dip to the line and turn back in the direction of the trend.
What breaks it: the trend ends and the level simply breaks. On markets that close overnight or at the weekend, a gap over news can skip the level entirely, so a stop beyond it fills at a worse price. That difference between the stop price and the fill is slippage, and it is most common when markets move fast.
What a false signal costs: the slippage on the stop, plus the round-trip spread or fee.
What to check first: whether the average still slopes in the direction of the trend, and whether news or a market close falls inside the time you expect to hold the trade.

Is there a best moving average for day trading?

No single length is best. Short lengths react faster and whipsaw more; long lengths whipsaw less and report turns later. On an intraday chart, a 9-period or 20-period EMA turns quickly, which also means it produces more signals, and every extra signal acted on is another round trip to pay for.

The useful question is not which average is best. It is how many signals a length produces in the market you trade, and whether the typical move after a signal is larger than its cost. A separate comparison of leading and lagging indicators sorts 12 tools by the same trade-off between speed and false signals.

FAQ: moving averages, answered

What is the 200-day moving average used for?

Mostly as a long-term trend filter on daily charts. Price above the 200-day average is read as a long-term uptrend, and price below it as a downtrend. Because it averages 200 closes, it moves slowly and turns late, so it describes the trend that has been, not the next one. In sideways markets, price crosses it repeatedly.

Why do moving averages lag?

Because every value is an average of past closes. When price turns, most of the closes in the window still belong to the old direction, so the average turns only after enough new closes replace them. The longer the window, the longer that takes. An EMA shortens the lag by weighting recent closes more, but cannot remove it.

Do moving averages work in a sideways market?

This is where they fail most. In a range, price crosses the average again and again, and crossovers reverse within days. Each cross can look like the start of a new trend, and each one that is traded is a round trip that pays the spread or fee. A flat average with price weaving around it is the warning sign.

What is a golden cross?

A golden cross is a shorter moving average, commonly the 50-day, crossing above a longer one, commonly the 200-day; a death cross is the same cross downwards. Both are crossovers, so both lag: they tend to appear after a trend change has already happened. A separate comparison covers the two in detail.

Conclusion

Moving averages describe the past. That is what makes them steady, and it is why each setup has a market that breaks it. The trend filter fails in sideways markets. The crossover fails on lag and in ranges. The pullback fails when the trend ends or price gaps through the level.

None of that means the lines are wrong. It means they are late by design, with a known failure and a known price for each failure. Before you rely on a setup, count how often it has flipped in the market you trade and what each flip costs to act on.

Trading leveraged products carries a high risk of loss. Check the spread and fee costs per market on Ouinex before you start counting.